Embed Size (px)
Citation preview
1
Fuel, Volume 181, 1 October 2016, Pages 522–530
Simulation of the calcination of a core-in-shell CuO/CaCO3 particle for Ca-Cu
chemical looping
Changlei Qin a,, Vasilije Manovic b,*, Jingyu Ran a, and Bo Feng c
a Key Laboratory of Low-grade Energy Utilization Technologies and Systems of Ministry of Education,
College of Power Engineering, Chongqing University, Chongqing 400044, China
b Combustion and CCS Centre, Cranfield University, Cranfield, Bedfordshire MK43 0AL, United
Kingdom
c School of Mechanical and Mining Engineering, The University of Queensland, St Lucia, Queensland
4072, Australia
Corresponding author:
Email: [email protected] (Changlei Qin); [email protected] (Vasilije Manovic)
2
ABSTRACT
The internal heat balance through heat generation due to CuO reduction and its consumption by CaCO3
decomposition makes calcination a critical step in a novel Ca-Cu chemical looping process (CaL-CLC).
Thus, the calcination behaviour of composite Ca/Cu particles needs to be well understood, especially
taking into account that mismatching of heat generation and consumption in the particles can lead to
local superheating, agglomeration and loss of activity due to enhanced sintering. In this work, a
composite particle model was developed to study the calcination behaviour within a spherical
core-in-shell type of particle containing grains of CuO and CaCO3. Simulation results showed that
ambient temperature, shell porosity, particle size, and CaCO3 grain size significantly affected the CuO
and CaCO3 reaction processes, while the impact of initial particle temperature and CuO grain size can be
ignored in the range of parameters considered in the study. By comparison of different types of particles,
it was concluded that the core-in-shell pattern was more advantageous if such particles are being applied
in CaL-CLC cycles due to better matching in reaction kinetics resulting in more stable and uniform
particle temperature distribution during the calcination stage.
Keywords: Ca-Cu chemical looping; CO2 capture; core-in-shell particle; composite particle model
3
1. INTRODUCTION
Global warming is becoming a severe problem attracting attention worldwide, due to the increasing
atmospheric concentrations of greenhouse gases (mainly CO2, CH4 and N2O). Among these greenhouse
gases, carbon dioxide accounts for half of the enhanced greenhouse effect and needs the most urgent
treatment [1]. CO2 capture and sequestration (CCS) has been identified as one of the most significant
ways to control CO2 emissions, and it could contribute 15-55% of the cumulative mitigation of
atmospheric CO2 by 2100 [2].
Calcium looping (CaL), particularly interesting for the cheap sorbents and high CO2 carrying capacity,
is a cost-effective method of capturing CO2 [3-5]. CaL has been rapidly developing, and the continuous
improvement of the process has resulted in its latest derivation, Ca-Cu chemical looping (CaL-CLC) [6,
7]. CaL-CLC could eliminate the energy intensively and costly air-separation unit in the conventional
CaL process by introducing a new chemical loop using CuO as the oxygen carrier, which could feed a
large amount of heat by its exothermic reduction for the highly endothermic regeneration of CaO-based
sorbents. Currently, the research on CaL-CLC is still in its early stage and only limited work has been
carried out on its development. Fernández et al. [8-10] reported a detailed conceptual design of Ca-Cu
chemical looping to obtain hydrogen and/or electricity from natural gas and a concentrated stream of
CO2 in fixed-bed reactors. Rahman et al. [11] discussed the integration of CaL-CLC with steam
gasification of biomass with the carbonator (gasifier), calciner, and air reactor arranged in three
sequences. Our team also worked in promoting the development of CaL-CLC, including the report of
the inhibition of CaO carbonation by CuO/Cu components [12], the mitigation methods [13], and the
assessment of kinetics matching on the calciner level [14].
It was well-known that the arrangement of CuO and CaCO3 in composite particles was an important
issue for the effective implementation of the CaL-CLC processes [9, 14]. Taking into account the heat
transfer efficiency and potential pressure drop inside the reactor, sorbents were usually prepared in the
form of spherical particles, with a combination of Ca-based and Cu-based materials, uniformly arranged
4
or in a core-in-shell pattern, or as separated particles [9, 15]. Kierzkoska and Müller [16] investigated
the synthesis of Ca-based, Cu-functionalised CO2 composites using a co-precipitation technique.
Manovic et al. [17, 18] prepared three types of pellets by granulation of powdered materials using a
mechanical pelletiser. The first type of pellet has a homogeneous distribution of CaO and CuO
supported by cement, and the other two types were both of the core-in-shell form but with different
arrangements of the active (CaO and CuO) and inert (cement) materials. They were shown to be suitable
for large-scale utilisation as no significant change in particle size was observed during a 2-h attrition test
in a bubbling fluidised bed. In contrast, Ridha et al. [15] reported a lower mechanical strength of
core-in-shell particles due to their susceptibility to fragmentation, in comparison to homogeneous
particles and mixed CaO-based and CuO-based particles.
It should be noted that the implementation of CaL-CLC largely depends on the effectiveness of
internal heat transfer from the reduction of CuO to the decomposition of CaCO3 within a single particle.
The synchronism of these two reactions is important and the coupling behaviour needs to be assessed.
Work on core-in-shell particles with CuO as core and CaCO3 as shell, which show potential in the
CaL-CLC application, has not been previously reported. Aiming to better understand the calcination
behaviour of core-in-shell particles during the calcination stage in CaL-CLC, a composite particle model
was developed to study the impact of operating conditions and particle parameters on the occurrence of
chemical reactions, gas species and heat profiles inside the particle. Furthermore, a comparison between
the performance of core-in-shell and uniformly-distributed particles was carried out based on the
simulation results.
2. COMPOSITE PARTICLE MODELING
The concept of the changing grain size model (CGSM) proposed by Georgakis et al. [19], has been
widely used in the literature [20-24]. In the present work, the model was further developed aiming to
understand the calcination behaviour of a core-in-shell particle with CuO and CaCO3 grains distributed
in the core and shell, respectively. A CuO/CaCO3 molar ratio of 3.2 was used in the particle simulation
5
considering the theoretical internal energy balance of chemical reactions. It was assumed that the
particle was porous and its radius remained constant during reactions, meaning that shrinkage of the
particle caused by sintering was omitted. The particle core also followed this assumption. The size of
the material grains changed as a result of the conversion of solid reactant to a product during the
reduction of CuO, as shown of Equation 1 (12.4 cm3/mol of CuO vs. 7.1 cm3/mol of Cu), and the
decomposition of CaCO3, as shown of Equation 2 (36.9 cm3/mol of CaCO3 vs. 16.9 cm3/mol of CaO).
Physical model of the core-in-shell particle and the change of grains during reactions are shown in
Figure 1.
4(g) (s) (s) 2(g) 2 (g) r,1173KCH 4CuO 4Cu CO 2H O H 207.6 kJ mol (1)
3(s) (s) 2(s) r ,1173KCaCO CaO CO H 165.6 kJ mol (2)
Figure 1. Schematic of the core-in-shell CuO/CaCO3 particle and the variation of a grain during
calcination.
For the specific core-in-shell particle, core radius is a function of the initial core and shell porosity and
the radius of the whole particle. Core radius is calculated as follows:
6
3
3, 0,0,
30 , 0, , 0,
3.2 1
3.2 1 1
m CuO shellcore
m CuO shell m CaCO core
VR
R V V
(3)
The mass balance for diffusion and reaction of gas ‘i’ inside the particle is given as follows:
2,2
1i ie i i
C CD R r
t R R R
(4)
The initial and boundary conditions required for solving the equation coupling external mass transfer
are as follows,
,0( , ) , 0i iC R t C t (5)
0
0, 0i
R
Ct
R
(6)
0
, , , , , 0ie i g i s i b i
R R
CD k C C t
R
(7)
The external mass transfer coefficient of ,g ik was calculated using the Sherwood number given by
[25],
0 1 2 1 32
2 0 6g ,i / /
m,i
k RSh . Re Sc
D (8)
The reaction rate of gas ‘i’ produced or consumed per unit of particle volume was proportional to the
chemical reaction rate constant k, and the following expressions with kinetic parameters for the
reduction of CuO with CH4 and the decomposition of CaCO3 [20, 26] were chosen for this work
because of their close fitting with the experimental data.
4
2
20
0
n,CuOCuO red ,CuO CH
,CuO
rr k S C
r
(9)
3 2
3 3
3 2
2
2
0
0
1,CaCO CO
CaCO dec ,CaCO
,CaCO eq ,CO
r Cr k S
r C
(10)
2
111 462 10 19130eq,CO
.C exp
T T
(11)
7
406000 5
redk . exp
gasR T(12)
131000254
deck exp
gasR T(13)
The initial specific surface of solid reactant ‘j’ was calculated as:
00
0
3 1, j v , j
, j
S fr
(14)
The radius variation of unreacted solid reactant was determined from the chemical reaction rate taking
place at the reaction interface with the equations:
4
2 n,CuOred m,CuO CH
drk V C
dt (15)
3 2
3
2
21
,CaCO CO
dec m,CaCO
eq ,CO
dr Ck V
dt C
(16)
The changing grain size of 1, jr was calculated as:
3 3 31 0 21, j j j , jr Z r Z r (17)
where jZ was defined as:
m,Pj
m,R
VZ
V (18)
The effective diffusivity of gas ‘i’ was calculated as a function of particle porosity as:
11 1 2
e ,i m,i K ,iD D D (19)
Particle porosity varies as a function of the initial porosity during reactions, and was calculated as:
0 0 01 1 v , j , j jf Z X ( R,T ) (20)
The molecular diffusivity of gas ‘i’ in the gas mixture was calculated by the Wilke equation:
8
1
1 im,i n
l
l ,l i m,( i ,l )
yD
y
D
(21)
where m,( i ,l )D expresses the molecular diffusivity of gas ‘i’ in gas ‘l’, and it was calculated using the
following equation [27]:
0 54 1 75 1 1
21 3 1 3
3 2 10..
i l
m,( i ,l )/ /
i l
. T M MD
P v v
(22)
The Knudsen diffusivity was calculated from:
6 135K ,i
e i
. TD
S M
(23)
The specific surface area of solid was calculated from the surface area of reactants:
2
1
01 0
n
, j
e , jj , j
rS S
r(24)
Local conversion inside the particle changes with time and location, and it was calculated as:
3
2
0
1, j
j
, j
rX R,t
r
(25)
The mean conversion of CuO and CaCO3 changing with time in the whole particle was calculated by
integration of local conversions:
0 2
0
30
4
4
3
,CuOR
j
CuO
,CuO
R X R,t dRX t
R
(26)
0
0
3
2
3 30 0
4
4
3
,CuO
R
jR
CaCO
,CuO
R X R,t dRX t
R R
(27)
For a spherical particle, the unsteady state heat transfer equation can be written as:
9
3
2
2
1p ef CaCO dec CuO red
T Tc R r H r H
t R R R
(28)
with the initial and boundary conditions being given as follows:
0 0T T , t (29)
0
0 0R
T, t
R
(30)
0
4 4 0
ef c b s b s
R R
Th T T e T T , t
R(31)
The heat convection coefficient in Equation (31) was estimated using the Nusselt number with the
correlation of Ranz and Marshall [28, 29] :
02
g
c
Nuh
R
(32)
0 5 0 332 0 0 6 . .Nu . . Re Pr (33)
The effective thermal conductivity of the particle depends on solid and gas thermal conductivities as
functions of porosity, and is calculated as follows:
1
1 2 11
3 3ef g s
g s
(34)
The thermal conductivity of the solid in a particle was calculated as a function of the volume fraction
of solid components in the particle, given as:
1s
v , j
s , j
f
(35)
The thermal conductivity of the gas mixture in a particle was calculated using the following equation:
1
1
ni g ,i
g ni
l ill
y
y A
(36)
with
10
20 5 0 25
0 5
0 5
1
8 1
. .
i i
l l
il .
. i
l
M
MA
M
M
(37)
The effective heat capacity was obtained by the following relation:
, ,1p s p s g p gc c c (38)
These equations were solved numerically on MATLAB v 7.9 by a finite volume method with a fully
implicit formulation and a time step of 10-4 s, and a convergence with a tolerance less than 10-6 was
chosen during the calculations in the work.
To verify the model, overall conversions of CuO and CaCO3 were obtained by calculation using the
developed model and from experiments in a thermogravimetric analyser. Considering the difficulty in
distinguishing the mass loss caused by CuO reduction and CaCO3 decomposition, the experiments were
carried out by testing the two reactions separately under isothermal conditions, and pure calcium
carbonate and CuO/alumina materials were utilised. During the testing, the total gas flow rate was
maintained at 100 mL/min, and conversions were calculated with the assumption that the mass loss was
only caused by the reactions studied. Figure 2 shows the comparison of the simulation and experimental
results, and it can be seen that the model developed properly predicts the rate of CuO reduction with
CH4 and the decomposition of CaCO3.
11
Figure 2. Comparison of simulation and experimental results of the conversion rate of (a) CaCO3
decomposition and (b) CuO reduction with CH4.
3. RESULTS AND DISCUSSION
To have a comprehensive view of the change in key variables with time, concentration profiles of CH4
and CO2, local temperature, porosity, local CuO and CaCO3 conversion, and grain size of CuO and
CaCO3 inside the spherical particle were displayed in a three-dimensional graph, as shown in Figure S1
in the supporting information. It was obtained by simulating a particle characterised with a porosity of
0.3 in both the core and shell, which were composed of CuO grains of 500 nm in radius and CaCO3
grains of 200 nm in radius, respectively. CuO and CaCO3 parameters used in the work were determined
according to the experimental data and calculation formulas presented in the literature [20, 26, 30-32].
The simulation was carried out for an ambient temperature of 1173 K and 0.2 bar of CH4 (balance CO2)
at atmospheric pressure with an initial particle temperature of 1173 K. Additionally, radial profiles of
variables at specific times under the same conditions as above are summarised in Figure 3.
12
Figure 3. Variation of radial distribution of some parameters within the particle with time (4CHP = 0.2
bar,2COP = 0.8 bar; bT = 1173 K, 0T = 1173 K;
30 0,CuO ,CaCO = 0.3, 0R = 1 mm, 0,CuOr = 500 nm,
30,CaCOr = 200 nm).
A sharp initial decrease in CH4 concentration from the particle surface to its centre can be seen in
Figure 3. At 5 s, CH4 concentration on the particle surface was 2 mol/m3 and it declined to around 0.2
mol/m3 on the core-shell boundary. Moreover, no CH4 was observed in the area of R/R0<0.62. This type
of CH4 distribution within the particle is a consequence of diffusion resistance in the CaCO3 shell owing
to the low shell porosity and its high effective specific surface area. Thereby, the reduction of CuO with
CH4 was very slow in the particle core area. In contrast, CO2 concentration was highest and almost
constant in the particle core, and it decreased from the core-shell boundary to the particle surface, which
13
was attributed to the combined effect of CO2 generation due to CaCO3 decomposition and CuO
reduction, and the varying local CO2 diffusivity in the particle.
Figure 3(c) shows that the temperature profile inside the particle was relatively flat over the studied
duration. It was constant at just above 1174 K in the core area and the value was a little lower on the
particle surface after 5 s, which was almost the same after 10 s. The profile of a slightly higher
temperature can be observed during the following 48 s without the appearance of a significant
temperature difference, implying that heat transfer in the particle and subsequent consumption due to
CaCO3 decomposition were well matched with CuO reduction. Then, particle temperature started to
decline slowly and reached a minimum value at around 68 s with the core temperature being slightly
lower than that in the shell, indicating the completion of CuO reduction and heat transfer from the
surroundings for the decomposition of CaCO3.
3.1. Effect of Ambient Temperature
The reduction of CuO and decomposition of CaCO3 are the two major reactions in the calcination
process of Ca-Cu chemical looping, and their kinetics are mainly determined by operating temperature
according to the Arrhenius law. Therefore, the influence of ambient temperature on the conversions was
investigated and is presented in Figure 4, for a particle of 1 mm in radius, with an identical initial
porosity of 0.3 in both the core (initial CuO grain radius of 500 nm) and the shell (initial CaCO3 grain
radius of 200 nm), and an initial particle temperature of 1173 K.
14
Figure 4. Profiles of temperatures and overall conversions inside the particle under different ambient
temperatures (4CHP = 0.2 bar,
2COP = 0.8 bar; 0T = 1173 K;30 0,CuO ,CaCO = 0.3, 0R = 1 mm, 0,CuOr =
500 nm,30,CaCOr = 200 nm).
It can be seen in Figure 4(a) that the overall conversion of CuO was very close to that of CaCO3
decomposition at an ambient temperature of 1173 K, although the latter proceeded slightly faster during
the first 30 s. Subsequently, the reduction of CuO gradually became slightly faster. The good coupling
of the two reactions, i.e., the almost simultaneous occurrence of CuO reduction and CaCO3
15
decomposition, can be a benefit for keeping the variation of particle temperature in a narrow range and
avoiding the appearance of superheating that can lead to sintering and agglomeration of materials, and
consequent loss of the activity and even operating life of the regenerator. Increasing the ambient
temperature to 1198 K, the CuO reduction rate was observed to increase more slowly than the rate of
CaCO3 decomposition. The reason for this was that while the former reaction was limited by the slow
feeding rate of CH4, the latter could still utilise the heat from the high-temperature surroundings. After
CaCO3 decomposition reached completion, a large amount of energy was released from the exothermic
reduction of the remaining CuO to heat the particle (the sharp temperature increase in Figure 4(b)). This
phenomenon became more apparent when ambient temperature was further increased to 1223 K. Thus,
1173 K was the optimal ambient temperature for the regeneration of sorbents under the conditions
studied.
3.2. Effect of Initial Particle Temperature
The initial temperature difference between sorbent particles and the surroundings requires a certain
time interval for the particles to reach the designated operating conditions. Thus, the effect of initial
particle temperature on conversion and temperature profiles was studied, ranging from 923 K to 1173 K,
while keeping the other variables constant. It can be seen from Figure 5 that the induction time for the
occurrence of CaCO3 decomposition became longer as decreasing initial particle temperature further
deviated from the ambient value, which agreed well with the work of Yin et al. [23] Though both CuO
reduction and CaCO3 decomposition became slower, the decline in the reaction rate of the latter was
more pronounced. In other words, the mismatching of reaction kinetics and heat transfer in the
regenerator could become more prominent by increasing the difference between initial particle
temperature and that of the surroundings.
16
Figure 5. Profiles of temperatures and overall conversions inside the particle while varying initial
particle temperature from 923 K to 1173 K (4CHP = 0.2 bar,
2COP = 0.8 bar; bT = 1173 K;
30 0,CuO ,CaCO = 0.3, 0R = 1 mm, 0,CuOr = 500 nm,30,CaCOr = 200 nm).
3.3. Effect of Particle Porosity
It is well known that porosity largely determines physical and chemical properties of a particle, such
as its crushing strength, attrition rate, inner gas diffusion, and reactivity [33], and the effect can be more
significant for core-in-shell particles because the feeding of CH4 from the surroundings to the core area
17
is restrained by the matrix of CaCO3 grains in the shell. To understand the potential impact, the
porosities of CuO in the core and CaCO3 in the shell were varied, respectively, from 0.1 to 0.5 and the
generated conversion profiles are presented in Figure 6. It can be seen that the overall conversion of
CuO rose with the increase of initial core porosity due to a lower resistance to the diffusion of CH4 into
the core region. As CuO reduction was the main heat source for CaCO3 calcination, the process of
CaCO3 decomposition changed accordingly and its variation was easily understood to be slightly
smaller than that of CuO. It was interesting to note from Figure 6(a) that coupling of the reactions of
CuO and CaCO3 changed only marginally with varying CuO porosity under the conditions studied.
Figure 6. Profiles of overall conversions inside the particle with varying initial porosity of (a) particle
core, and (b) particle shell (4CHP = 0.2 bar,
2COP = 0.8 bar; bT = 1173 K, 0T = 1173 K; 0R = 1 mm,
0,CuOr = 500 nm,30,CaCOr = 200 nm).
Figure 6(b) shows the profile of overall conversions as a function of initial shell porosity. It was
known that the existence of a shell containing CaCO3 grains was the main barrier for the penetration of
CH4 from the ambient to the particle core. The decrease of initial CaCO3 porosity caused a reduction in
its specific surface area, which together led to a sharp fall-off of effective diffusivity of CH4 in the shell.
As a result, overall conversion of CuO was observed to decline quickly. With regard to CaCO3
18
decomposition, it was also largely inhibited owing to the lack of heat supply from CuO reduction. This
condition remained unchanged until much of the CaCO3 was converted, at which point the rate of CuO
reduction with CH4 increased quickly. When increasing the initial porosity of CaCO3 to 0.5, its
inhibition on CH4 diffusion was weakened. Thus CuO reduction became much faster, and the
calcination of CaCO3 was enhanced. However, process matching of the two reactions was slightly
poorer at a shell porosity of 0.5 than that with a value of 0.3.
3.4. Effect of Particle Size
Particle size usually affects reaction rates and external mass and heat transfer, which has been
observed in the single reduction of CuO and CaCO3 decomposition [20, 33]. Here, we investigated the
core-in-shell CuO/CaCO3 composite by varying particle radius from 0.5 mm to 1.5 mm, and
summarised the results in Figure 7. It was apparent that the conversion of both CuO and CaCO3
decreased with increasing particle radius. For the particle with a radius of 1 mm, CuO and CaCO3
achieved full conversion in 66 s and 117 s, respectively, which were extended to 135 s and 146 s at a
particle radius of 1.5 mm. The heat released from CuO reduction could still be used in calcining CaCO3,
but the efficiency varied according to the reaction process. As can be seen for the particle with a radius
of 0.5 mm, the rate of CuO reduction was much faster than CaCO3 decomposition, thus a large amount
of energy was released over a short time and the whole particle was heated up. After all CuO was
consumed, some of the CaCO3 remained unreacted and its further reaction would depend on the heat
transferred from the surroundings, which resulted in a lower particle temperature than that of the
ambient. For the 1.5-mm-radius particle, the reaction process of CuO was slightly behind CaCO3
decomposition for most of the time; therefore, the particle temperature was below the ambient
temperature. It was clear that once the reaction rate of CuO exceeded that of CaCO3, the heat released
was sufficient to support CaCO3 decomposition and particle temperature would rise to higher than that
of the surroundings, as shown in Figure 7(b).
19
Figure 7. Profiles of overall conversions and temperatures inside the particle with varying particle
radius (4CHP = 0.2 bar,
2COP = 0.8 bar; bT = 1173 K; 0T = 1173 K;30 0,CuO ,CaCO = 0.3, 0,CuOr = 500
nm,30,CaCOr = 200 nm).
3.5. Effect of Grain Size
The potential effect of grain size variation on calcination behaviour was investigated by changing
20
grain radius of CuO and CaCO3 in the range of 200 nm to 800 nm, and the results are summarised in
Figure 8. It can be seen that overall conversion profiles of CuO were almost identical, regardless of the
variation of CuO grain size. The distribution of CaCO3 conversion with time was also maintained.
These observations mean that CuO grain size in the range investigated has almost no effect on the
simultaneous reaction behaviour of CuO and CaCO3 during regeneration. The reason for this is that the
reduction of CuO is kinetically fast under the conditions studied, and only the supply of CH4 from the
surroundings via the shell could be a limitation. However, the diffusion of CH4 was unaffected when
varying grain size of CuO in the core area. As a result, the decomposition process of CaCO3 remained
unchanged as well. By contrast, the rate of CuO reduction became faster with increased initial CaCO3
grain radius, which slightly accelerated CaCO3 decomposition initially, as shown in Figure 8(b). As
CuO reduction finished, the decomposition rate of CaCO3 in the case of a larger initial grain size
decreased more quickly and the time required to reach 100% conversion was prolonged.
Figure 8. Profiles of overall conversions inside the particle with varying initial grain radius of (a) CuO,
and (b) CaCO3 (4CHP = 0.2 bar,
2COP = 0.8 bar; bT = 1173 K; 0T = 1173 K;30 0,CuO ,CaCO = 0.3, 0R =
1 mm).
21
3.6. Comparison of Uniformly-distributed and Core-in-shell Particles
Uniform-distribution and core-in-shell patterns are the two most interesting arrangements for CuO
and CaCO3 grains in a composite particle that can potentially be used in the CaL-CLC system due to
good matching in heat release and the subsequent utilisation. In our previous work [34], the time to
reach full conversion of CuO and CaCO3 was calculated to be 35 s and 137 s, respectively, when a
particle of 1 mm in radius and 0.3 in porosity with a uniform distribution of CuO and CaCO3 grains
(grain size of 500 nm vs. 200 nm) was initialised at 1173 K under 0.2 bar of CH4. Under the same
conditions, the time for reaction completion varied between 66 s and 117 s for the core-in-shell particle.
Additional simulation showed that including or neglecting heat radiation between the particle and its
surroundings in equation (31) makes no difference for the conditions compared. The time extension for
CuO to reach full reduction was clearly attributable to the resistance of the CaCO3 shell to CH4
diffusion as analysed above. However, its inhibition of CuO reduction could accelerate the occurrence
of CaCO3 decomposition and resulted in a more synchronised calcination/reduction, i.e., a better
matching of heat release and utilisation. This coupling can be pursued in CaL-CLC system design, as it
eliminates the requirement of an external heat source for regenerating CaCO3, which is one inherent
advantage of CaL-CLC compared to the conventional calcium looping for CO2 capture. Furthermore,
good matching of CuO reduction and CaCO3 decomposition rates could help to avoid large particle
temperature fluctuations with time. This conclusion can be easily observed from the temperature
variation range of 55 K for the uniformly-distributed particle and only around 10 K for the core-in-shell
type, which is an advantage in preventing superheating and stabilising operation of the regenerator.
Therefore, the core-in-shell arrangement was more suitable for application in the calcination stage of
CaL-CLC if taking reaction kinetics and heat transfer within particles into consideration.
4. CONCLUSIONS
To well understand the behaviour of reaction coupling (CuO reduction with CH4 and the
22
decomposition of CaCO3) and heat transfer in the calcination stage of the novel Ca-Cu chemical looping
process, a mathematical model of the composite particle was developed to investigate the core-in-shell
arrangement within the framework of the key particle parameters and operating conditions. Simulation
results showed that ambient temperature, porosity of the shell, particle size, and CaCO3 grain size
significantly affected the reaction process of CuO and CaCO3. A lower ambient temperature, in the
range of 1173-1223 K, was observed to be beneficial for coupling of the reactions, while an increase of
ambient temperature resulted in faster decomposition of CaCO3 when compared to that of reduction of
CuO. Though the coupling behaviour of the two reactions was similar with the change of core porosity
from 0.1 to 0.5, a significant decline of the reduction rate of CuO was observed when decreasing the
shell porosity from 0.3 to 0.1 as a result of the high resistance of CH4 diffusion in the shell area. It was
also found that the optimal radius of the core-in-shell particle should be around 1 mm or a little larger
with a small CaCO3 grain (200 nm) under the conditions studied in this work. In contrast, the effect of
CuO grain size and the initial particle temperature can be neglected.
Compared to particles with a uniform distribution of CuO and CaCO3 grains, the core-in-shell
arrangement was determined to be beneficial for coupling the CuO reduction with CH4 and
decomposition of CaCO3, with the particle temperature fluctuating only in a narrow range. The model
developed can be a useful tool for quantifying trade-offs between the key design and operational
parameters of the calcination process in CaL-CLC.
NOTATION
4CHC CH4 concentration within the particle, mol/m3
2COC CO2 concentration within the particle, mol/m3
2eq ,COC equilibrium CO2 concentration, mol/m3
iC concentration of gas ‘i’, mol/m3
23
0i,C initial concentration of gas ‘i’ within the particle, mol/m3
b,iC concentration of gas ‘i’ in the reactor, mol/m3
s,iC concentration of gas ‘i’ on particle surface, mol/m3
,p gc specific heat of gas mixture, J/(kgK)
,p sc specific heat of solid within particle, J/(kgK)
e ,iD effective diffusivity of gas ‘i’, m2/s
K ,iD Knudsen diffusion coefficient of gas ‘i’, m2/s
m,iD molecular diffusion coefficient of gas ‘i’, m2/s
e emissivity
v , jf volume fraction of solid ‘j’ within the particle
0v , j ,f initial volume fraction of solid ‘j’ within the particle
gasR gas constant, J/(molK)
ch external heat transfer coefficient, W/(m2K)
g ,ik external mass transfer coefficient of gas ‘i’, m/s
redk reaction rate constant of CuO reduction, m1.3875/(mol0.4625s)
deck reaction rate constant of CaCO3 decomposition, m3/(mols)
iM molecular weight of gas ‘i’, kg/mol
lM molecular weight of gas ‘l’, kg/mol
Nu Nusselt number
n reaction order of the reduction of CuO with CH4
P total pressure, Pa
Pr Prandtl number
24
ir production rate of gas ‘i’, mol/(m3s)
CuOr rate of CuO reduction, mol/(m3s)
3CaCOr rate of CaCO3 decomposition, mol/(m3s)
0 , jr initial grain radius of solid ‘j’, m
1, jr grain radius of solid ‘j’ when reaction proceed, m
2 , jr grain radius of unreacted solid ‘j’, m
0,CuOr initial grain radius of CuO, m
2,CuOr grain radius of unreacted CuO, m
30 ,CaCOr initial grain radius of CaCO3, m
32 ,CaCOr grain radius of unreacted CaCO3, m
R radial coordinate, m
0R initial radius of the particle, m
Re Reynolds number
0,CuOS initial specific surface area of CuO, m2/m3
30 ,CaCOS initial specific surface area of CaCO3, m2/m3
0 , jS initial specific surface area of solid reactant ‘j’, m2/m3
eS specific surface area of the solid, m2/m3
Sc Schmidt number
Sh Sherwood number
t time, s
T temperature, K
0T initial particle temperature, K
25
bT temperature in the reactor, K
sT temperature of the particle surface, K
m,CuOV molar volume of CuO, m3/mol
3m ,CaCOV molar volume of CaCO3, m3/mol
m,PV molar volume of solid product, m3/mol
m,RV molar volume of solid reactant, m3/mol
jX conversion of solid reactant ‘j’
iy molar fraction of gas ‘i’
jZ volume ratio for solid product to reactant
Greek symbols
0 initial particle porosity
particle porosity
g thermal conductivity of the gas mixture, W/(mK)
ef effective thermal conductivity within the particle, W/(mK)
s thermal conductivity of the solid within the particle, W/(mK)
,s j thermal conductivity of solid ‘j’, W/(mK)
redH heat of CuO reduction, J/mol
decH heat of CaCO3 decomposition, J/mol
i dynamic viscosity of gas ‘i’, kg/(ms)
l dynamic viscosity of gas ‘l’, kg/(ms)
g density of gas mixture, kg/m3
26
s density of solid within the particle, kg/m3
i
v diffusion volume for molecule ‘i’
Stefan-Boltzman coefficient, W/(m2K4)
ACKNOWLEDGEMENT
The authors are grateful to the financial supports from China Postdoctoral Science Foundation (No.
2015M572448), Chongqing Postdoctoral Funding Project (No. Xm2015002), the Fundamental Research
Funds for the Central Universities (No. 106112015CDJXY140005), Key Laboratory of Low-grade
Energy Utilization Technologies and Systems of Ministry of Education (No. LLEUTS-201411,
LLEUTS-2016004), and the Scientific Research Foundation for the Returned Overseas Chinese
Scholars, State Education Ministry.
REFERENCES
[1] IPCC. IPCC Fourth Assessment Report: Climate Change 2007-Working Group I: The Physical
Science 2007.
[2] IPCC. IPCC Special Report on Carbon Dioxide Capture and Storage. Prepared by Working Group
III of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New
York, NY, USA: Cambridge University Press; 2005.
[3] Li Z, Liu Y, Cai N. Understanding the enhancement effect of high-temperature steam on the
carbonation reaction of CaO with CO2. Fuel. 2014;127:88-93.
[4] Materić V, Symonds R, Lu D, Holt R, Manović V. Performance of Hydration Reactivated Ca
Looping Sorbents in a Pilot-Scale, Oxy-fired Dual Fluid Bed Unit. Energy Fuels.
2014;28:5363-72.
[5] Itskos G, Grammelis P, Scala F, Pawlak-Kruczek H, Coppola A, Salatino P, et al. A comparative
characterization study of Ca-looping natural sorbents. Appl Energy. 2013;108:373-82.
27
[6] Abanades JC, Murillo R, Fernandez JR, Grasa G, Martínez I. New CO2 Capture Process for
Hydrogen Production Combining Ca and Cu Chemical Loops. Environ Sci Technol.
2010;44:6901-4.
[7] Feng B. Sorbent regeneration. 2011;PCT/AU2011/000007, Australia
[8] Fernandez JR, Abanades JC, Grasa G. Modeling of sorption enhanced steam methane
reforming—Part II: Simulation within a novel Ca/Cu chemical loop process for hydrogen
production. Chem Eng Sci. 2012;84:12-20.
[9] Fernández JR, Abanades JC, Murillo R, Grasa G. Conceptual design of a hydrogen production
process from natural gas with CO2 capture using a Ca–Cu chemical loop. Int J Greenhouse Gas
Control. 2012;6:126-41.
[10] Martínez I, Romano MC, Fernández JR, Chiesa P, Murillo R, Abanades JC. Process design of a
hydrogen production plant from natural gas with CO2 capture based on a novel Ca/Cu chemical
loop. Appl Energy. 2014;114:192-208.
[11] Rahman RA, Mehrani P, Lu DY, Anthony EJ, Macchi A. Investigating the Use of CaO/CuO
Sorbents for in Situ CO2 Capture in a Biomass Gasifier. Energy Fuels. 2015.
[12] Qin C, Yin J, Liu W, An H, Feng B. Behavior of CaO/CuO Based Composite in a Combined
Calcium and Copper Chemical Looping Process. Ind Eng Chem Res. 2012;51:12274-81.
[13] Qin C, Yin J, Luo C, An H, Liu W, Feng B. Enhancing the performance of CaO/CuO based
composite for CO2 capture in a combined Ca–Cu chemical looping process. Chem Eng J.
2013;228:75-86.
[14] Qin C, Feng B, Yin J, Ran J, Zhang L, Manovic V. Matching of kinetics of CaCO3 decomposition
and CuO reduction with CH4 in Ca–Cu chemical looping. Chem Eng J. 2015;262:665-75.
[15] Ridha FN, Lu D, Macchi A, Hughes RW. Combined calcium looping and chemical looping
combustion cycles with CaO–CuO pellets in a fixed bed reactor. Fuel. 2015;153:202-9.
[16] Kierzkoska A, Müller CR. Development of calcium-based, copper-functionalised CO2 sorbents to
28
integrate chemical looping combustion into calcium looping. Energy Environ Sci. 2012;5:6061-5.
[17] Manovic V, Anthony EJ. Integration of calcium and chemical looping combustion using
composite CaO/CuO-based materials. Environ Sci Technol. 2011;45:10750-6.
[18] Manovic V, Wu Y, He I, Anthony EJ. Core-in-Shell CaO/CuO-Based Composite for CO2 Capture.
Ind Eng Chem Res. 2011;50:12384-91.
[19] Georgakis C, Chang CW, Szekely J. A changing grain size model for gas—solid reactions. Chem
Eng Sci. 1979;34:1072-5.
[20] García-Labiano F, Abad A, de Diego LF, Gayán P, Adánez J. Calcination of calcium-based
sorbents at pressure in a broad range of CO2 concentrations. Chem Eng Sci. 2002;57:2381-93.
[21] Mahuli SK, Agnihotr R, Jadhav R, Chauk S, Fan LS. Combined calcination, sintering and
sulfation model for CaCO3-SO2 reaction. AICHE J. 1999;45:367-82.
[22] Yu YS, Liu WQ, An H, Yang FS, Wang GX, Feng B, et al. Modeling of the carbonation behavior
of a calcium based sorbent for CO2 capture. Int J Greenhouse Gas Control. 2012;10:510-9.
[23] Yin J, Kang X, Qin C, Feng B, Veeraragavan A, Saulov D. Modeling of CaCO3 decomposition
under CO2/H2O atmosphere in calcium looping processes. Fuel Process Technol.
2014;125:125-38.
[24] Sedghkerdar MH, Mahinpey N. A Modified Grain Model in Studying the CO2 Capture Process
with a Calcium-Based Sorbent: A Semianalytical Approach. Ind Eng Chem Res. 2015;54:869-77.
[25] Schlichting H, Gersten K. Boundary-Layer Theory. 8th ed. Berlin & New York: Springer; 2000.
[26] Juan Adánez, Francisco García-Labiano, Luis F. de Diego, Ainhoa Plata, Javier Celaya, Pilar
Gayán, et al. Optimizing the Fuel Reactor for Chemical Looping Combustion. 17th International
Conference on Fluidized Bed Combustion. Jacksonville, Florida, USA, 2003. p. 173-82.
[27] Fuller EN, Schettler PD, Giddings JC. New method for prediction of binary gas-phase diffusion
coefficients. Ind. Eng. Chem. 1966;58:18-27.
[28] Ranz WE, Marshall WR. Evaporation from drops, Part I. Chem Eng Prog. 1952;48:141-6.
29
[29] Ranz WE, Marshall WR. Evaporation from drops, part II. Chem Eng Prog. 1952;48:173-80.
[30] Borgwardt RH. Calcination kinetics and surface area of dispersed limestone particles. AICHE J.
1985;31:103-11.
[31] García-Lario AL, Martínez I, Murillo R, Grasa G, Fernández JR, Abanades JC. Reduction
Kinetics of a High Load Cu-based Pellet Suitable for Ca/Cu Chemical Loops. Ind Eng Chem Res.
2013;52:1481-90.
[32] Davidson ML, Harrison DP. Evaluation of the changing grain size model. Chem Eng Commun.
1984;25:213-28.
[33] García-Labiano F, de Diego LF, Adánez J, Abad A, Gayán P. Temperature variations in the
oxygen carrier particles during their reduction and oxidation in a chemical-looping combustion
system. Chem Eng Sci. 2005;60:851-62.
[34] Qin C, Yin J, Feng B, Ran J, Zhang L, Manovic V. Modelling of the calcination behaviour of a
uniformly-distributed CuO/CaCO3 particle in Ca–Cu chemical looping. Appl Energy.
2016;164:400-10.
